The Design Optimization Of Condenser Based On Numerical Simulation

As one of the most important components for nuclear power system, condenser has a crucial influence on the weight and the overall arrangement of nuclear power system. In the previous studies on condenser optimization, when calculating the heat transfer rate in the evaluation function, most researchers incorporate the empirical method (e.g. HEI standard),which is simply and time-saving. However, with the increase of condenser power and the complexity of bundle arrangement, empirical method might be no longer suitable. To solve this problem and improve the practicability of the optimization procedure, the application of numerical modeling in the genetic algorithm optimization method has been tested in the pre-sent study.A numerical two-dimensional condenser model consist of four differential equations is present here, as well as the discretization of equations, definition of boundary conditions and the selection of suitable solver. The numerical calculation procedure is coded in Python with Numpy,a mathematical extension,and then applied to the modeling of the prototype conden-ser. The result has been displayed graphically and analyzed, which confirmed the rationality of the program.The contractive mapping genetic algorithm has been chosen for the condenser weight optimization. To take the influence of different operators on the optimization efficiency into consideration, different combinations of mutation operators, crossover operators and selection operators are examined and compared. It is showed that the combination of non-uniform mu-tation operator, heuristic crossover operator and roulette selection operator has better effi-ciency compared with the other combinations.The optimization procedure with numerical two-dimensional condenser model and con-tractive mapping genetic algorithm has been tested. To solve the time-consuming evaluation function, the master-salve genetic algorithm is applied. The result confirmed the feasibility of applying numerical modeling in condenser optimization procedure. The difference of optimi-zation with numerical modeling and that with empirical method is also compared,in both the optimization results and optimization curves. The numerical model can take account of more parameters, which brings a wider application range to the optimization procedure. Besides,the condensate rate with certain cooling tube number is chosen as optimal objective, in an at-tempt to optimize the condenser efficiency, with venting and buddle arrangement as optimal variables respectively,and the results show different levels of improvement in efficiency.